We are interested in the mechanisms of cell signaling and innate immunity. In particular, we are fascinated by how cells detect microbial infections and other noxious agents and how they mount an appropriate response to eliminate infectious agents and dangerous molecules without harming the host.
A common feature of all microbes, including viruses, bacteria, fungi, and parasites, is that they contain DNA and/or RNA in their genome. The detection of these microbial nucleic acids is a highly effective and general mechanism of host defense that is evolutionarily conserved. However, because host cells also contain abundant DNA and RNA, it is crucial that the host immune system avoids attacking its own nucleic acids. In humans, inappropriate reactions to self–nucleic acids are a major culprit of autoimmune diseases such as lupus. On the other hand, our immune system has also evolved to take advantage of the nucleic acid–sensing mechanism to detect not only microbes but also some of our own cells that have gone bad, such as cancer cells. This mechanism of cancer detection is critical for our intrinsic antitumor immunity.
Microbial infections deliver their nucleic acids into the host endosomes, cytoplasm, and/or nuclei. Certain Toll-like receptors are localized on the endosomal membrane to detect the nucleic acids in the lumen of the endosome and transmit a signal to cytoplasmic proteins that ultimately lead to the production of antimicrobial and immune modulatory molecules such as type I interferons (e.g., IFNα and IFNβ) and inflammatory cytokines. For productive infections, microbial DNA and RNA must also gain access to and replicate in the host cytosol and/or nuclei. How animal cells detect cytosolic DNA and RNA to mount an immune response is an active area of research in our lab.
RIG-I Pathway of Cytosolic RNA Sensing
RNA viruses, such as influenza and hepatitis C, produce viral RNAs when they enter and replicate inside the host cells. These viral RNAs are detected by the cytosolic protein RIG-I and/or its related proteins MDA5 and LGP2, collectively known as RIG-I–like receptors (RLRs). The RIG-I ligands are duplex RNAs containing 5'-triphosphate or 5'-diphosphate; such signatures are absent in host cytoplasmic RNA that normally contains 5' modifications such as 5' cap. The binding of viral RNA to the C terminus of RIG-I induces a conformational change that exposes the N-terminal tandem CARD domains (2CARD), which mediate downstream signaling to induce interferons and other cytokines.
We have found that the RIG-I 2CARD binds to lysine-63–linked polyubiquitin (K63 polyUb) chains, which are not conjugated to other cellular proteins. The binding of these unanchored K63 polyUb chains causes RIG-I 2CARD to form a tetramer, which then interacts with the CARD domain of an adapter protein that we named MAVS (mitochondrial antiviral signaling). The interaction between the RIG-I 2CARD tetramer and MAVS CARD causes the latter to form a self-perpetuating filament that functionally resembles a prion. In this process, the RIG-I 2CARD tetramer "nucleates" MAVS CARD, which then polymerizes to form the functional filaments. These filaments recruit cytosolic TRAF proteins, including TRAF2, TRAF5, and TRAF6, which we have previously shown to function as ubiquitin E3 ligases that promote polyubiquitination to activate the cytosolic protein kinases IKK and TBK1. IKK and TBK1 in turn activate the transcription factors NF-κB and IRF3, respectively, which enter the nucleus to regulate the expression of a plethora of genes encoding immune and inflammatory mediators.
Using a cell-free system, we have reconstituted the RIG-I pathway from RNA ligands to the activation of IRF3 and NF-κB. We are using this system to further dissect the biochemical mechanism of signal transduction in the RIG-I pathway. We are also investigating the mechanism by which the assembly of the prion-like filaments is regulated. This line of investigation recently led to the discovery that the inflammasome adapter protein ASC also forms prion-like filaments, which are important for activating the inflammatory signaling cascade.
cGAS Pathway of Cytosolic DNA Sensing
With the exception of RNA viruses, which are recognized by the RLR pathway, all other microorganisms contain DNA or require DNA in their life cycles. Living organisms ranging from bacteria to humans have evolved sophisticated cell-intrinsic mechanisms of immune defense against microbial DNA (e.g., the CRISPR system in bacteria). In animal cells, where the cellular DNA is normally in the nucleus or mitochondrion, a major mechanism of host defense is to detect the presence of foreign DNA in the cytoplasm.
Through a biochemical approach, we recently identified the long-sought cytosolic DNA sensor that triggers the interferon pathway. This sensor turns out to be a novel enzyme, which we named cyclic GMP-AMP (cGAMP) synthase (cGAS), that binds to DNA and becomes activated to synthesize cGAMP from GTP and ATP. cGAMP then functions as a second messenger that binds to and activates the adapter protein STING, which in turn activates the kinases IKK and TBK1 to induce interferons and cytokines.
We have solved the crystal structures of cGAS in its apo- and DNA-bound forms, which reveal DNA-induced dimerization and rearrangement of the active site that lead to the activation of the enzyme (Figure 1). We also solved the structures of cGAMP and its complex with STING. The endogenous cGAMP molecule contains mixed phosphodiester linkages. This cGAMP isoform, which we termed 2'3'-cGAMP, binds to STING with a high affinity; this binding induces a conformational change of STING, resulting in its activation.
To investigate the role of cGAS in vivo, we generated a cGAS-knockout mouse strain. The cGAS deficiency abolishes the induction of interferons and other cytokines by DNA transfection or DNA virus infection. The cGAS-knockout mice are more susceptible than wild-type mice to lethality caused by DNA virus infections. We also showed that cGAS is an innate immune sensor of retroviruses, including HIV, by detecting the reverse-transcribed viral DNA. These studies validate cGAS as a nonredundant cytosolic DNA sensor.
The discovery of the cGAS pathway sets the stage for dissection of the signaling mechanisms in this pathway. Furthermore, because aberrant activation of the cGAS pathway may cause some human autoimmune diseases, such as lupus, chemical inhibitors of cGAS or STING may be developed for the treatment of these diseases. On the other hand, we have shown that cGAMP is a potent adjuvant that boosts antibody production and T cell activation in mice. Thus, cGAMP and its derivatives may be further developed as adjuvants for vaccines and immunotherapy.
This research is supported in part by grants from the National Institutes of Health, the Cancer Prevention and Research Institute of Texas (CPRIT), and the Robert Welch Foundation.
As of March 17, 2016